This project involves both computational and experimental aspects. Visual search is a primary function of visual behavior. Little is known about the underlying mechanisms, however, particularly in understanding how the visual search process is manifest in eye movements, especially in a natural task. The experiments examine the eye movements made during search for a target object in a naturalistic scene. Our experiments reveal unexpected but very regular patterns of fixations. Gaze was typically directed to the center of the image before being shifted to the target, as shown in Figure ??. In this situation, `serial' search does not involve sequential fixation of individual objects, and is much more easily explained in terms of signal/noise models and a coarse-to-fine spatial analysis. In a separate theoretical effort we have been able to predict the pattern of fixations in human subjects using a simple model based on biologically plausible computations such as those performed in the primary visual cortex. A target for search can be simply described by the response of oriented spatial filters at different spatial scales (as might be carried out by simple cells in Vi). Rao and Ballard have shown that such a code discriminates well between different regions in a visual display, and can be used to reliably relocate the region in a visual search task (Figure ??). A simple filter template like this could be used for the target selection process, and used by the observer to mark locations to be held in memory to guide subsequent saccades. The linking of the human search experiments and computational modeling in this tight fashion is one of the major scientific achievements of the Resource. We will continue this work in the context of studies with patients with localized lesions to further strengthen the modeling and link it more closely with the underlying brain mechanisms.